The Logic Behind Innovation: How Ozelle Is Reshaping Diagnostics with AI + CBM

The Starting Point — From Traditional CBC Analyzer to Intelligent Blood Diagnostics

For decades, the Complete Blood Count (CBC) has been the cornerstone of hematology diagnostics, delivering essential quantitative data — red blood cells, white blood cells, hemoglobin, platelets — but telling only part of the story.

Clinical decisions often depend on what lies beyond the numbers: the morphological characteristics of blood cells that reveal early pathological changes — subtle shape variations, cytoplasmic granularity, nuclear irregularities.

Today, these insights still rely on manual microscopy performed by highly trained specialists — a time-consuming, subjective, and non-scalable process. This diagnostic gap became Ozelle’s starting point:

Can we teach an analyzer not only to count cells, but to understand them?

This question gave rise to Ozelle’s AI+CBM (AI-driven Complete Blood Morphology) program — a multi-year research initiative to bridge hematology with artificial intelligence, optics, and computational biology.

The Exploration — Advancing Blood Cell Count Analysis with AI + CBM and POC Innovation

Ozelle began developing Complete Blood Morphology with a clear vision: to standardize, quantify, and automate morphological interpretation through deep learning, enabling analyzers to “see” cells with the precision of a hematologist.

Key Technological Breakthroughs: Inside Ozelle’s AI-Powered, Multi-Functional Hematology Analyzer

• High-Definition Cell Imaging Pipeline A customized optical module captures multi-dimensional cell images with positional repeatability of <1 μm, ensuring high-fidelity morphological preservation and consistency across measurements.
• AI Recognition Engine Trained on over 40 million expert-annotated clinical cell images, Ozelle’s deep-learning model classifies diverse cell types — including RET, NSG, NSH, ALY, PAg, RET and various abnormal cells — with expert-level accuracy. Recognized at the World Artificial Intelligence Conference (WAIC) 2022, the model continuously improves through self-learning as more data are acquired.
• Morpho-Quantitative Fusion Model Integrating quantitative CBC parameters with AI-based morphology analysis, this approach generates comprehensive diagnostic profiles that combine cell counts with structural interpretation.
• Edge-AI Architecture A fully automated workflow — from sample loading to result generation — completes in almost 6 minutes per sample, ensuring operational independence, data security, and high-throughput efficiency.

These innovations converge in a compact analyzer that seamlessly performs CBC + AI Morphology within a standard laboratory routine.

From R&D to Real-World Blood Testing Devices — Bridging Laboratory Precision and POC Efficiency

Transforming the AI+CBM concept into a manufacturable system required deep co-engineering across optics, algorithms, and biomedical fluidics.

• Optical Engineering：Ozelle’s imaging channel achieves higher cellular resolution and complete morphological visualization compared with conventional impedance-based systems.
• AI Pipeline: The adaptive classification algorithm filters, segments, and identifies cell populations in real time — automatically flagging anomalies for expert review and replacing manual differential counts in most routine cases.
• Data Validation: Multi-center clinical evaluations conducted across the Asia-Pacific, Latin America, and Middle East regions demonstrate that AI+CBM can reduce manual review rates, while maintaining diagnostic accuracy and reproducibility.
• Simple to Use and Maintain: One-click operation, single-use reagent kits, and no sample pre-treatment. Designed for maximum efficiency, minimal maintenance, and room-temperature reagent storage — redefining usability in everyday diagnostics.

These capabilities now define Ozelle’s flagship product line — designed for decentralized testing networks such as healthcare facilities, clinics, and pharmacies, as well as research laboratories and veterinary diagnostic applications.

Market Impact — How Multi-Functional Hematology Analyzers Are Redefining Blood Diagnostics

AI-driven Complete Blood Morphology reshapes both the operational and business value chains of diagnostics.

For distributors and integrators, Ozelle’s innovative systems create a new category of diagnostic solution — compact, intelligent, and commercially scalable:

• Expand portfolios into AI-powered diagnostics with low entry barriers.
• Differentiate product lines in mature markets where CBC has become commoditized.
• Unlock multi-vertical potential across human, veterinary, and research domains.
• Reduce service burdens with maintenance-free hardware and cloud-upgradable software.

In short, Ozelle transforms morphological complexity into operational simplicity — converting diagnostic data into sustainable business value.

The Vision Ahead — Building an AI-Driven Diagnostics Ecosystem for Humans, Clinics, and Veterinary Care

Ozelle’s long-term strategy extends far beyond instrument innovation.

By embedding AI-driven morphological analysis into compact, multi-functional analyzers, we are building the foundation of an intelligent diagnostics ecosystem — one that connects humans, clinics, and veterinary care under a shared AI framework.

This ecosystem unifies hematology, biochemistry, and immunology data to deliver diagnostics that are:

• Smarter: driven by AI, refined through continuous data learning.
• Faster: integrated across multiple disciplines for real-time insights.
• More Connected: powered by cloud collaboration and remote consultation.

At Ozelle, innovation begins with a question and ends with transformation — from data to diagnosis, from automation to intelligence, and from instruments to ecosystems.